Tuberculosis (TB) is a treatable infectious disease, however, the lack of efficient diagnostic tests are important bottleneck impeding access to treatment. Thus, simple, sensitive and specific diagnostic approaches are desperately needed. An antigen detection assay for TB is uniquely attractive because tests can be performed using a variety of patient specimens. The urine lipoarabinomannan- enzyme-linked immunosorbent assay (LAM-ELISA) detections are attractive, especially for the patients with HIV infection, unfortunately, the sensitivity of these assays is low, which makes this test suboptimal. The question we ask is why the sensitivity of LAM-ELISA detection is inefficient? LAM from Mtb grown in vivo may have a structure quite unique and distinct from its in vitro counterpart and it could have distinct modifications such as succinylation, mannosylation and substitution with methythioxylofuranose which will lend credence to its sensitivity. If so, it will explain the failure in reaction with antibodies included in the commercial kits to capture LAM in biofluid as all are raised against in vitro grown Mtb or LAM isolated from Mtb grown in vitro. Thus in this proposal we propose to exploit our unique ability to extract LAM from lung, granulomas, liver and kidneys from experimentally infected guinea pigs with different clinical strains, and to define the nonreducing end exhaustively focusing on the terminal residue modifications since they are the epitopes for the anti-LAM antibodies recognition. Initial proof of concept experiments has shown that the yield of LAM in vivo will not be sufficient for the structural analysis using the methods for the structural analysis on LAM from the in vitro grown bacilli. Therefore, our structural studies will rely heavily on use of more sensitive tests such as Gas Chromatography-Mass Spectrometry (GC-MS), enzyme digestion, and mapping of the oligosaccharides with Electrospray Ionization. As a fingerprint, in vitro grown bacilli from clinical isolates with definite phenotype will be included in the protocol. The definition of the terminal ends of LAM in vivo will impact the future when a sensitive and specific anti-LAM antibody can be generated based on this study to enhance LAM based existing assays for use in resource poor regions.